Growing up, I’ve always been fascinated by TV shows and films portraying robots, machines, and computers as having intelligence that rivals some of the smartest people to ever exist and performing tasks that seem unthinkable for a non-living thing. Imagine solving equations in the blink of an eye, constructing images from nothing but your imagination, or even searching every source available for detailed information. Impossible, right?
Over recent years, artificial intelligence has been improving dramatically and shows signs of increasing further to a point that seemed impossible before but is soon-to-be achievable. With on-going research and development in these fields, it begs the question: what does a future with AI look like, and what is the full extent that it can reach? In order to have a better understanding of what our future may look like in terms of AI, let us first delve into the world of artificial intelligence at its current level.
The Wonders of Artificial Intelligence
AI refers to a wide range of computers and programs that have the ability to conduct human-like tasks. Using mathematics and programming, AI is capable of problem solving, decision-making, and adapting to unique situations. AI engineers around the world working alongside scientists have been able to create AI systems and programs that help businesses in security, earning profits of more than 5% with AI alone (Global Survey: The State of AI in 2021, 2021) and improving efficiency for businesses by more than 40% (Dziedzic, 2023). Altogether, AI has changed the way the world operates and made working much simpler, more timely, and more cost-efficient.
A common misconception is that artificial intelligence only refers to one area when, in reality, it is split into various diverse fields that are further broken down. It can usually be classified as weak AI or strong AI. The difference between the two is that weak AI is only capable of performing a specific task it was designed to do and lacks in other areas (What Is Strong AI?, n.d.), while strong AI is capable of replicating humans and performing actions on its own accord. Ever heard of Siri, ChatGPT, or self-driving Teslas? These popular examples of weak AI, along with many others, are the types that are commonly used in our era for efficiency, entertainment, and making menial labour much easier. In comparison, strong AI is typically showcased in films and only theoretically exists, like in WALL-E and the Terminator. The lack of strong AI may not be a bad thing, however, because the reality of a computer with the same (or greater) knowledge as humans, emotions, and ability to think on its own could be disastrous.
Artificial intelligence also has other sectors, like machine learning and computer vision. Machine learning is the branch of artificial intelligence that allows machines to predict outcomes and come to conclusions by analyzing patterns and viewing data (What Is Machine Learning?, n.d.). In different terms, it’s a way for AI to learn without humans being too involved in teaching it. Machine learning can be further split into branches like deep learning and neural networks, which allow AI to process higher-level data similar to the brain (Deep Learning Vs. Machine Learning: Beginner's Guide, 2023). Computer vision, on the other hand, is another branch of artificial intelligence that analyzes images in order to compare data, and it works by a machine teaching itself by using machine learning in order to view many images and discern key features and differences (What Is Computer Vision?, n.d.). By doing so, the machine is able to distinguish what it is viewing, which is similar to the AI having eyes of its own.
The All-Inclusive World of AI
With the global artificial intelligence market expected to increase to two trillion US dollars within the next decade compared to the one hundred billion it is right now, the hope of further advancements in AI is high (Thormundsson, 2023). This massive increase in value means that within the next few years, more companies, businesses, and customers will invest in and purchase AI products to suit their needs. The increase in popularity of AI comes from the fact that it is suitable for nearly every industry and is extremely useful and viable. The most prominent uses of AI are in healthcare, customer service, and social media networks. However, recently, a form of AI known as generative AI has reached record-breaking levels for being known to constantly develop creative applications and produce high-scale outputs.
In healthcare, AI is used to diagnose patients, most commonly by identifying irregularities within the body like tumours, using computer vision methods. It is also able to recognize diseases and diagnose patients by analyzing patient histories, searching lab data, and viewing medical imaging. Recently, hospitals around the world have started implementing robotic arms that are capable of assisting or conducting specific types of surgery. This type of aid goes beyond what a surgeon could be able to do alone since it reduces the risks of making mistakes and leads to precise cutting techniques (Daley, n.d.).
In customer service, AI has been applied to assist customers online by answering questions and offering the full services of company websites. Aside from personalized responses, AI in customer service offers non-stop support and constantly improves its responses by utilizing the concept of machine learning. Social media companies like Instagram, Youtube, Twitter, and Facebook have also all used AI in their apps by showing users posts that are suitable to them based on the history of what they were interested in viewing on their feed.
A specific type of AI known as generative AI has caused an uproar in the past few months for multiple reasons. If you had the opportunity to finish a task that would otherwise take you days or weeks, would you take it? This type of AI is capable of creating media, text, and images on its own that would otherwise take a person much longer to complete. These include programs that write essays, create art and imagery given text input, complete pieces of code given statements, and even create musical masterpieces all in a matter of seconds. With the introduction of ChatGPT by OpenAI earlier last year, many companies have started producing generative AI models that can compete with DALL-E, Bard, and Bing AI, and as a result, there has been a spike in demand for artificial intelligence worldwide.
An interesting study showcased in March by two Japanese researchers showed that an AI algorithm called Stable Diffusion had the capability of reading brain scans and producing images as a result (Bove, 2023). Using fMRI brain scans of data from the occipital and temporal lobes that record information about images that a viewer looks at, as well as key words and text that were translated from the data, the AI algorithm was able to produce images that closely resembled the original. Using only the raw neural activity information, the AI matched the layout of the image, and with the aid of text, it could use its trained database to add details to the final product (Nahas, 2023). This form of generative AI might only be applied minimally currently; however, with more testing and improvements, it has the potential to aid in communication for people who cannot speak directly and turn memories into physical images.
The applications of artificial intelligence in our world are constantly evolving, and it’s exciting to find out new ways in which AI can be used in our lives. From healthcare to reading minds, our current level of artificial intelligence is vast and most definitely impressive. But what does a future with further developed AI look like, and is it even possible to achieve such a stage?
An Insight Into The Future
Alan Turing, also known as the father of computer science, once said, "A computer would deserve to be called intelligent if it could deceive a human into believing that it was human." The Turing test is a method that can be used to determine if a machine has intelligence indistinguishable from that of a human. It operates with a human operator communicating with another human and a machine, with each participant separated from each other. The operator asks several questions, and if they are unable to tell which participant is a machine, then the AI is said to have passed (Turing Test, n.d.). While it may be true that we have come a long way in AI development since the first ever design in 1955, there have been no AI models or algorithms that have passed this test yet.
Turing’s words hold a lot of value in the sense that no AI has reached a point at which it would be perfectly comparable to humans, and although artificial intelligence may seem limitless at first, there are many things to consider before achieving a higher stage regarding AI.
With the advancement of AI comes a risk to privacy because this form of technology collects a lot of data and can use it without permission at any time (Iftikhar, 2021). An example of this is facial recognition, which scans a person’s face and saves it for future use. If the data is stolen, it can easily be misused for identity theft, and malicious actions done with a person’s face would be difficult to invalidate. Concerns with privacy limit AI’s potential in the future since algorithms run on data to continuously improve results and operate effectively. With some countries, like China and the majority of Europe, banning this form of data collection, further improvements to AI would prove difficult.
According to a survey by CNBC, more than 24% of workers believe that their occupation will be overridden by AI in the future (Rosenbaum, 2023). Despite some public disapproval, it is important to note that further development of AI will create new jobs to replace old ones, with an expected 133 million jobs created compared to 75 million eliminated by 2025, according to a study by the World Economic Forum (The Future of Jobs Report 2023 | World Economic Forum, 2023). By training algorithms to assess past data and study patterns using neural networks (a branch of ML), it is possible for AI to make quick decisions and deliver tasks that repeat themselves. Positions like data analysts and data entry, customer service, media management, and warehouse operators are at the most risk of losing their positions to AI, but positions like AI engineers, AI trainers, data scientists, and robotics engineers are said to grow (Roos, 2023). In the future, it is more likely that employees will work alongside AI to complete tasks quickly rather than allowing AI to manage positions on their own due to limitations and the risk of error.
Robots and computers planning to take over the world or fall in love with a person extremely exaggerate the capabilities that AI can achieve, at least at the current time. At the moment, AI is unable to express emotional feelings or think on its own because it is impossible to replicate the human brain. Emotions come from evolution, sensing surroundings and interactions, having internal thoughts, and personal experiences from the past (Rudin, 2019). Without giving AI consciousness or the ability to think on its own with a brain, it is virtually impossible for it to have its own feelings. At most, by scanning neural data from fMRI imaging of a person looking at an image, an AI algorithm could determine what kinds of emotions the person felt when viewing it (Rudin, 2019). Until the barrier between programming an AI to follow commands and allowing it to think on its own is broken, the future of AI is limited to learning by analyzing patterns and absorbing data and can only improve when people feed it more data.
With further ideas needed for new methods AI can use to learn, gain the ability to think, and give responses comparable to a person, there is a long way to go in this field before a truly perfect AI algorithm or model is born. We can expect AI to live and work alongside us in ways like building, helping the ill, handling transportation, and lessening work loads. Beyond the limitations of current AI lies a mysterious trove of innovative wonders, and we can expect AI to continue surprising us each day!
By introducing new and innovative designs from engineers and researchers, it won’t be shocking to see how and where AI can be applied next. From chatbots to smart cars to personalized recommendations to banking, artificial intelligence is everywhere! Continuing to advance, AI will be the key to a highly technological world where the new age of computers, machines, and robots can flourish and make our lives easier.
Bove, T. (2023, March 9). New A.I. model replicates images based on brain scans. Fortune. Retrieved August 2, 2023, from https://fortune.com/2023/03/09/ai-generate-images-human-thoughts-stable-diffusion-study/
Daley, S. (n.d.). AI in Healthcare & Medical AI Examples to Know. Built In. Retrieved August 1, 2023, from https://builtin.com/artificial-intelligence/artificial-intelligence-healthcare
Deep Learning vs. Machine Learning: Beginner's Guide. (2023, June 15). Coursera. Retrieved August 3, 2023, from https://www.coursera.org/articles/ai-vs-deep-learning-vs-machine-learning-beginners-guide
Dziedzic, M. (2023, February 22). AI in the Workplace: The Future of Business Ef iciency and Cost Savings. LinkedIn. Retrieved August 1, 2023, from https://www.linkedin.com/pulse/ai-workplace-future-business-efficiency-cost-savings-daniel-calciano
The Future of Jobs Report 2023 | World Economic Forum. (2023, April 30). The Future of Jobs Report 2023 | World Economic Forum. Retrieved August 3, 2023, from https://www.weforum.org/reports/the-future-of-jobs-report-2023
Global survey: The state of AI in 2021. (2021, December 8). McKinsey. Retrieved August 1, 2023, from https://www.mckinsey.com/capabilities/quantumblack/our-insights/global-survey-the-state-of-ai-i n-2021
Iftikhar, N. (2021, September 19). Is Artificial Intelligence a Threat to Privacy? ThinkML. Retrieved August 2, 2023, from https://thinkml.ai/is-artificial-intelligence-a-threat-to-privacy/
Nahas, K. (2023, March 7). AI re-creates what people see by reading their brain scans. Science.org. Retrieved August 2, 2023, from https://www.science.org/content/article/ai-re-creates-what-people-see-reading-their-brain-scans
Roos, D. (2023, June 6). 8 Jobs AI Will Replace and 8 It Won't (Yet). Electronics | HowStuffWorks. Retrieved August 3, 2023, from https://electronics.howstuffworks.com/future-tech/jobs-ai-will-replace.htm
Rosenbaum, E. (2023, June 8). These are the American workers most worried that A.I. will soon make their jobs obsolete. CNBC. Retrieved August 3, 2023, from https://www.cnbc.com/2023/06/08/these-are-the-workers-most-worried-that-ai-will-soon-take-the ir-jobs.html
Rudin, P. (2019, October 18). AI and the Issue of Human Emotions › SINGULARITY 2030. Singularity 2030. Retrieved August 3, 2023, from https://singularity2030.ch/ai-and-the-issue-of-human-emotions/
Thormundsson, B. (2023, July 17). Artificial Intelligence market size 2030. Statista. Retrieved July 30, 2023, from https://www.statista.com/statistics/1365145/artificial-intelligence-market-size/
Turing test. (n.d.). Wikipedia. Retrieved August 2, 2023, from https://en.wikipedia.org/wiki/Turing_test
What is Computer Vision? (n.d.). IBM. Retrieved August 1, 2023, from https://www.ibm.com/topics/computer-vision
What is Machine Learning? (n.d.). IBM. Retrieved July 31, 2023, from https://www.ibm.com/topics/machine-learning
What is Strong AI? (n.d.). IBM. Retrieved July 31, 2023, from https://www.ibm.com/topics/strong-ai